Model Explainer

Feature Importances

Feature Importances

Model Summary

metric Score
mean-squared-error 23515.885
root-mean-squared-error 153.349
mean-absolute-error 110.378
mean-absolute-percentage-error 2125055810202728.2
R-squared 0.735

Predicted vs Actual

Residuals

Plot vs feature

Are predictions and residuals correlated with features?

Individual Predictions

Select Index

Selected index: 39225

Prediction

totalRent
Predicted 452.180
Observed 424.000
Residual -28.180

Contributions Plot

How has each feature contributed to the prediction?

Partial Dependence Plot

Contributions Table

How has each feature contributed to the prediction?
Reason Effect
Average of population 676.86
livingSpace = 38.0 -190.55
numberOfFloors_more_than_5 = 1.0 +47.73
serviceCharge = 100.0 -34.25
kitchen = 0.0 -26.54
noRooms = 1.0 +22.72
cellar = 0.0 +17.85
petsAllowed_no = 0.0 -14.62
interiorQuality_not_luxury = 0.0 -13.65
interiorQuality_normal = 1.0 -13.35
balcony = 0.0 -8.78
garden = 0.0 +7.09
lastRefurbish_cat_very new = 0.0 -6.83
newlyConst = 0.0 -6.1
interiorQuality_luxury = 0.0 -2.59
telekomUploadSpeed_cat_slow = 0.0 +2.34
interiorQuality_nan = 0.0 -2.17
numberOfFloors_4 = 0.0 -2.15
lastRefurbish_cat_old = 0.0 +1.44
telekomUploadSpeed_cat_fast = 1.0 -0.99
energyEfficiencyClass_C_OR_LOWER = 1.0 -0.98
telekomTvOffer_ONE_YEAR_FREE = 1.0 +0.8
telekomUploadSpeed_cat_nan = 0.0 -0.75
typeOfFlat_non_luxury_type = 1.0 -0.64
energyEfficiencyClass_A = 0.0 -0.59
lastRefurbish_cat_nan = 1.0 +0.57
numberOfFloors_nan = 0.0 +0.37
numberOfFloors_3 = 0.0 +0.36
numberOfFloors_2 = 0.0 -0.32
telekomTvOffer_nan = 0.0 +0.25
typeOfFlat_penthouse = 0.0 -0.23
typeOfFlat_terraced_flat = 0.0 -0.19
lastRefurbish_cat_new = 0.0 -0.17
numberOfFloors_0 = 0.0 +0.11
energyEfficiencyClass_B = 0.0 +0.09
typeOfFlat_maisonette = 0.0 +0.06
numberOfFloors_1 = 0.0 -0.05
telekomTvOffer_ON_DEMAND = 0.0 +0.05
typeOfFlat_loft = 0.0 +0.01
lastRefurbish_cat_very old = 0.0 +0.0
petsAllowed_yes = 1.0 +0.0
Other features combined +0.0
Final prediction 452.18

What if...

Select Index

Selected index: 95481

Prediction

totalRent
Predicted 578.834

Feature Input

Adjust the feature values to change the prediction

Contributions Plot

How has each feature contributed to the prediction?

Partial Dependence Plot

Contributions Table

How has each feature contributed to the prediction?
Reason Effect
Average of population 676.86
livingSpace = 81.78 +105.02
serviceCharge = 171.65 -13.63
kitchen = 0.0 -31.68
interiorQuality_not_luxury = 0.0 -17.04
balcony = 0.0 -39.16
newlyConst = 0.0 -16.59
petsAllowed_no = 0.0 -25.28
interiorQuality_normal = 0.0 +9.83
noRooms = 4.0 -30.6
numberOfFloors_more_than_5 = 0.0 -2.62
lastRefurbish_cat_very new = 0.0 -10.99
garden = 1.0 -16.02
telekomUploadSpeed_cat_slow = 1.0 -6.32
interiorQuality_luxury = 0.0 -2.6
cellar = 1.0 -2.6
energyEfficiencyClass_C_OR_LOWER = 1.0 -1.63
lastRefurbish_cat_old = 0.0 +1.09
interiorQuality_nan = 1.0 -3.62
numberOfFloors_4 = 0.0 +2.25
numberOfFloors_nan = 0.0 -1.3
typeOfFlat_non_luxury_type = 1.0 -1.82
telekomTvOffer_ONE_YEAR_FREE = 1.0 +1.79
telekomUploadSpeed_cat_nan = 0.0 -1.6
energyEfficiencyClass_A = 0.0 -0.54
lastRefurbish_cat_nan = 1.0 -0.12
telekomUploadSpeed_cat_fast = 0.0 +6.83
numberOfFloors_2 = 0.0 -1.0
numberOfFloors_3 = 1.0 +1.25
telekomTvOffer_nan = 0.0 +0.3
lastRefurbish_cat_new = 0.0 -0.09
typeOfFlat_penthouse = 0.0 -0.51
numberOfFloors_1 = 0.0 +0.15
energyEfficiencyClass_B = 0.0 +0.54
numberOfFloors_0 = 0.0 +0.34
telekomTvOffer_ON_DEMAND = 0.0 +0.2
typeOfFlat_maisonette = 0.0 +0.21
typeOfFlat_terraced_flat = 0.0 -0.37
typeOfFlat_loft = 0.0 -0.08
lastRefurbish_cat_very old = 0.0 +0.0
petsAllowed_yes = 1.0 +0.0
Other features combined +0.0
Final prediction 578.83

Feature Dependence

Shap Summary

Ordering features by shap value

Shap Dependence

Relationship between feature value and SHAP value

Feature Interactions

Interactions Summary

Ordering features by shap interaction value

Interaction Dependence

Relation between feature value and shap interaction value

Decision Trees

Select Index

Selected index: 139

Decision Trees

Displaying individual decision trees inside xgboost model

Decision path table

Decision path through decision tree
no tree selected